Abstract: MATH/CHEM/COMP 2002, Dubrovnik, June 24-29, 2002



Initial analysis of binding sites in protein-DNA interaction based on affinity of hydrogen bonds

and van der Waals contact formation


Anita KriSko1, Larisa ZoraniC2, and Michael Gromiha3


1Rudjer Boskovic Institute, POB 180, HR-10002 Zagreb, Croatia


2Department of Physics, University of Split, HR-21000 Split, Croatia


3RIKEN TSUKUBA Institute, Tsukuba, Japan



Vital cellular processes like DNA replication, genetic expression, DNA repair and packing engage binding of proteins to DNA. Recently, many 3D structures of protein-DNA complexes have been elucidated and nature of interaction has been examined. The simplest way to study these interactions is by using binding preferences of amino acids and nucleotides. With respect to the DNA recognition motifs, 62 nonredundant sequences (out of 129 solved tertiary structures of protein-DNA complexes) were classified into eight DNA-binding protein structural families1:

(i) helix-turn-helix proteins (21),

(ii) zinc-coordinating proteins (11),

(iii) zipper-type proteins (4),

(iv) other α-helical proteins (4),

(v) β-sheet proteins (3),

(vi) the β-hairpin/ribbon proteins (6),

(vii) enzymes (4),

(viii) and others (9).

Each group of sequences was aligned using CLUSTAL-X. Experimentally solved binding
regions for zinc-coordinating, zipper-type, other α-helical and the β-hairpin/ribbon proteins were highlighted in multiple sequence alignments.

Based on observed frequency of hydrogen bonding, van der Waals contacts and water mediated hydrogen bonding reported by Luscombe et al.2 binding preferences of amino acid-nucleotide interactions were calculated. Preference profiles were made for all 62 sequences. Binding regions in zipper-type proteins are correlated with regions of highest preference values in preference profiles. Binding domains of other α-helical proteins have similar profiles: central amino acid(s) with high preference(s) and surrounding amino acids with low or average preferences. The β-hairpin/ribbon protein binding regions (except 1CMA_A) have asymmetric profiles: section with lower preferences followed by section with higher amino acid preferences. For all other classes an overall rule for binding regions was not found.


1 N. M. Luscombe, S. E. Austin, H. M. Berman, J. M. Thornton, An overview of the structures of protein-DNA complexes, Genome Biology 1 (2000) 1-10.

2 N. M. Luscombe, R. A. Laskowski, J. M. Thornton, Amino acid-base interactions: a three-dimensional analysis of protein-DNA interactions at an atomic level, Nucleic Acids Research 29 (2001) 2860-2874.